基于python的邮件地址提取小程序

import sys

import os

import re



def analysis_file(path):

    print("analysis file: %s." % path)

    fi = open(path, "r")

    try:

        all_text = fi.read()

    finally:

        fi.close()



    #print("content:")

    #print(all_text)

    

    mails = set()

    re_mail = re.compile(r"([a-zA-Z-]+(?:\.[\w-]+)*@[\w-]+(?:\.[a-zA-Z-]+)+)")



    ms = re_mail.findall(all_text)

    for m in ms:

        #print(m)

        mails.add(m)



    print("results: %d" % len(mails))



    if len(mails) > 0:

        fo = open(path + ".mail.txt", "wt")

        for mail in mails:

            fo.write(mail)

            fo.write(",")

        fo.close()



def analysis_dir(path):

    files = os.listdir(path)

    for file in files:

        if (not os.path.isfile(file)) or file.endswith(".mail.txt"):

            continue

        analysis_file(path + "\\" + file)



def main():

    print("analysis is working... ...")

    print("current direcotry: %s." % os.getcwd())

    if len(sys.argv) < 2:

        print("set the directory to serach")

        return

    path = sys.argv[1]

    is_file = os.path.isfile(path)

    if is_file:

        print("searching file: %s." % path)

        analysis_file(path)

    else:

        if not os.path.exists(path):

            print("there isn't exist direcoty: %s" % path)

            return

        print("searching alll files in directory: %s." % path)

        analysis_dir(path)



if __name__ == '__main__':

    main()

这是我在实际过程中常碰到的问题，大量的邮件地址分散在若干文件中，可以处理单个文件或目录，提取所有的邮件地址，并拼接成邮件地址列表，可以直接由于批量邮件的发送。